作者：舒征宇1, 沈佶源1, 李黄强2, 熊会林2, 李世春1, 马聚超1
作者单位：1. 三峡大学电气与新能源学院，湖北 宜昌 443002;
2. 国网湖北省电力有限公司宜昌供电公司，湖北 宜昌 443000
关键词：覆冰厚度;扫描线种子填充算法;纹理特征;fcm算法
摘要：
针对覆冰厚度监测过程中常伴随雨雪天气，严重影响图像质量与覆冰厚度辨识精度的问题，文中提出一种基于航拍图像的输电线路覆冰厚度辨识方法。首先采用改进的扫描线种子填充算法对采集所得的输电线路图像中的雪花进行填充，以减少雪花对图像质量的影响；其次对线路覆冰的纹理特征进行加权构成融合特征，并结合空间邻域信息的fcm算法，实现对线路区域的提取；最后利用输电线路实际直径与覆冰前后线路上下边界像素宽度确定覆冰厚度。实验结果表明，文中所提方法能够有效辨识雨雪天气下的输电线路覆冰厚度，计算所得的覆冰厚度相对误差仅为1.05%，对线路除冰工作具有一定的参考价值。
research on the identification of transmission line ice thickness based on aerial images
shu zhengyu1, shen jiyuan1, li huangqiang2, xiong huilin2, li shichun1, ma juchao1
1. college of electrical engineering & new energy, china three gorges university, yichang 443002, china;
2. yichang power supply company of state grid hubei electric power co., yichang 443000, china
abstract: to address the problem that the monitoring process of ice cover thickness is often accompanied by rain and snow, which seriously affects the image quality and ice cover thickness recognition accuracy, an aerial image-based transmission line ice cover thickness recognition method is proposed in the paper. firstly, the improved scanning line seed filling algorithm is used to fill the snowflakes in the acquired transmission line images to reduce the impact of snowflakes on the image quality. secondly, the texture features of the line ice cover are weighted to form fusion features and combined with the fcm algorithm of spatial neighborhood information to realize the extraction of the line area. finally, the actual diameter of the transmission line and the pixel width of the upper and lower boundaries of the line before and after the ice cover are used to determine. the actual diameter of the transmission line and the pixel width of the upper and lower boundaries of the line before and after the ice cover are used to determine the ice cover thickness. the experimental results show that the proposed method can effectively identify the transmission line ice thickness under rain and snow, and the relative error of the calculated ice thickness is only 1.05%, which has certain reference value for line deicing work.
keywords: ice cover thickness;scanline seed filling algorithm;texture feature;fcm algorithm
2023, 49(4):21-25,59 收稿日期: 2022-09-11;收到修改稿日期: 2022-10-24
基金项目: 国家自然科学基金项目(51907104)
作者简介: 舒征宇(1983-),男,湖北武汉市人,副教授,博士,主要从事智能电网运维、含新能源的电力系统优化规划
参考文献
[1] 陈立帆, 张琳琳, 宋辉, 等. 基于图卷积神经网络的输电线路自然灾害事故预测[j/ol]. 电网技术: 1-9[2022-07-30].
[2] 黄欢, 杨旗, 吴建蓉, 等. 架空地线电磁脉冲除冰装置参数优化[j/ol]. 中国测试: 1-7[2022-10-21].
[3] sun w, wang c f. staged icing forecasting of power transmission lines based on icing cycle and improved extreme learning machine[j]. journal of cleaner production, 2018, 208(pt.1-1658): 1384-1392
[4] weng b j, gao w, zheng w c, et al. newly designed identifying method for ice thickness on high-voltage transmission lines via machine vision[j]. high voltage, 2021, 6(5): 904-922
[5] ma y, yu h, liu j y, et al. measurement of ice thickness based on binocular vision camera[c]// 2017 ieee international conference on mechatronics and automation (icma). ieee, 2017.
[6] zhang y y, wang y l, wei a m. a new image detection method of transmission line icing thickness[c]// 2020 ieee 4th information technology, networking, electronic and automation control conference (itnec). ieee, 2020.
[7] 王伟斌, 李立学, 李海峰, 等. 基于小波变换和形态学处理的输电线路覆冰监测系统[j]. 电气自动化, 2020, 42(4): 63-65
[8] 肖文, 高宏力, 鲁彩江. 单目视觉测量电力线覆冰厚度方法研究[j]. 机械设计与制造, 2021(8): 1-4
[9] 姚景岩, 杨世勇, 舒征宇, 等. 基于航拍的输电线路覆冰厚度监测研究[j]. 高压电器, 2021, 57(8): 105-114
[10] 闫丽梅, 褚思琦, 徐建军, 等. 基于gso-canny算法的输电线路覆冰厚度检测[j]. 国外电子测量技术, 2022, 41(2): 157-161
[11] 林刚, 王波, 彭辉, 等. 基于强泛化卷积神经网络的输电线路图像覆冰厚度辨识[j]. 中国电机工程学报, 2018, 38(11): 3393-3401
[12] 孙杰. 基于android平台图像处理算法的研究与实现[d]. 北京: 北京邮电大学, 2011.
[13] 温银堂, 张松, 张玉燕, 等. 新型复合材料界面粘接缺陷的ct检测及表征[j]. 中国测试, 2020, 46(1): 12-17
[14] 陈旭, 高亚洲, 陈守静, 等. 基于t-glcm和tamura融合特征的纹理材质分类[j/ol]. 南京信息工程大学学报(自然科学版): 1-11[2022-07-13].
[15] 朱浩, 丁辉, 尚媛园, 等. 多纹理分级融合的织物缺陷检测算法[j]. 纺织学报, 2019, 40(6): 118-125
[16] 岑若晨, 李建良. 改进pso-2d otsu在连铸坯缺陷图像分割中的应用[j]. 中国测试, 2020, 46(4): 19-24
[17] lai j w, zhu h q, ling x f. segmentation of brain mr images by using fully convolutional network and gaussian mixture model with spatial constraints[j]. mathematical problems in engineering, 2019(11): 1-14
